At the moment, we are all being bombarded with numbers associated with New Labour’s financial legacy. Well there is another number which reveals a further legacy of the last administration.
Under New Labour, for each UK citizen in the last year (i.e. you), there was a 1 in 50 chance that your communications could have been accessed by the law enforcement and national security authorities. This translates into a “Governmental Annual Surveillance Probability” (known as the “GASP”) of at least 0.02.
How do I get to that? From the Annual Report of the Interception of Communications Commissioner for 2009 (published last July), which I have just had the pleasure of reading.
Anyway, let’s get down to some revealing statistics from the Report.
The Report notes that there were 1,500 warrants signed by a relevant Cabinet Minister who authorised interception of communications, and that were 5,267 subsequent “modifications” to these warrants over the year. This means that each warrant is changed on average 3.5 times and that each warrant has four or five different forms: the original warrant (which the Cabinet Minister has to authorise) and three or four subsequent modified forms (that the Cabinet Minister is unlikely to see). The Scottish figures are more or less the same: 192 warrants in force and 629 changes (3.3 modifications per warrant).
The nature of these modifications are not mentioned in the Report. They could be trivial (e.g. adding a new phone number used by the target of the surveillance) - but who knows? One can only hope that the “modifications to warrants” are not substantial, for if they were to be substantial, then a “modification procedure” is being used as a substitute for the procedure that should apply when a new warrant is being obtained.
If this were to be the case, the matter becomes very serious as any protection afforded by Regulation of Investigatory Powers Act (RIPA) with respect to warrant authorisation procedures is undermined. And that is why it is important for commentary on the number of modifications to warrants to appear in the Annual Report. None does – so we have to rely on “hope”.
The Report also notes that 192 warrants were signed by the Scottish Government; this relates to policing matters which is a devolved matter. If we assume that the rate of warrants per head of population for the police is a constant (i.e. the Scottish population of 5 million is as criminal as the “rest of the UK” population of 56 million), then the “rest of the UK” should have about 2,150 warrants for policing alone.
However, the actual total number of reported warrants for the rest of the UK is 1,514 and this covers all policing plus prisons and all the national security agencies. This means that either the rest of the UK need far fewer warrants, or the Scots need more warrants, or that in “the rest of the UK” warrants are drafted differently from the Scottish equivalents. For instance, if the rest of the UK warrants were far more widely drawn with respect to authorised communications interception than the Scottish ones, then obviously you need fewer of them (especially as there are 3.5 modifications to come).
Given that the Report covers the period when New Labour Ministers were at the helm of surveillance authorisation in Westminster, and an opposition SNP Minister was authorising in Edinburgh, it is not difficult to guess which of the three options above is the more likely.
If this hypothesis is correct, the 600+ figure describing the discrepancy in the number of warrants signed by Ministers measures New Labour Ministers’ toleration of broadly defined surveillance practices.
Note an implication for future Annual Reports. If there were to be more warrants signed by Coalition Ministers reported, then this does not automatically mean there has been more interception. It could mean that surveillance practices have been more precisely defined before being authorised and more warrants are needed to cover the same range of interception.
Whatever the reason, the numbers appear to reveal a major difference in operational practice with respect to warrant authorisation by Scottish and Westminster Ministers that needs an explanation. One should have been given in the Annual Report as it is very easy to see the issue from the statistics.
Now to the current political debate about Local Authority surveillance. The Report indicates that there were 525,130 requests for communications data of which 1,756 requests were made by Local Authorities. Hence to remove Local Authorities from access to communications data (0.33% of all communications data requests) will not change much.
So any journalist, privacy activist or politician who argues that Local Authorities should not have to access communications data and that this is “a major privacy gain”, is in effect, deluding himself. This is because 99.6% of all access to communications data are excluded from his analysis.
The Report also indicates that other ad-hoc bodies account for a further 2,259 requests. If this is added to the Local Authority requests and subtracted from the total number of requests, we are drawn to the conclusion that approximately 520,000 requests for communications data are made by the police and security services per year. Now each request could cover a large number of individuals or telephones (this is explained in blog of 9/9/2009 – see references), but we can at least assume that each request has to cover at least 2 individuals (the caller and the called).
If we assume that each request for communications data involves at least two people over the age of 15, then the 520,000 requests involve at least 1.04 million adults. As there are 51 million adults in the UK, we can then evaluate that about 1 adult in 50 has been subject to this form of communications surveillance by the police or security services per year.
This translates into a New Labour GASP of 0.02 and the true number is probably much higher! The 1 in 50 number is quantifiable as New Labour’s surveillance legacy and it also benchmarks how the new Coalition Government approaches the issue. One should add that it is much larger than the GASP reported by the Daily Mail on its front page last September (see references).
Whatever the arguments about numbers, it is clear that the Interception of Communications Commissioner should have been able to explain something to reassure the public. He should recognise that it is not the number of requests that should be reported, it is the number of individuals who have been subject to communications data surveillance or the average number of individuals affected per request. Only then will the true scale of state surveillance will become apparent.
However, I leave a final statistic to the reader. How many marks out of 10 do you give an Annual Report that is intended to reassure the public that RIPA works properly, but which misses all of the above?
More about the Governmental Annual Surveillance Probability (or GASP).
As the GASP is a probability it has to be a number between 0 and 1. When GASP=0, then there is no communications surveillance, but if GASP = 1 (this is known by privacy activists as “the last GASP”) there is total surveillance. Surveillance at the family level (say 1 in 10) is GASP = 0.1. GASP = 0.04 (1 in 25) would be the probability that one of your friends would be under communications surveillance. So it follows that New Labour’s GASP = 0.02 shows a high degree of surveillance.
My comment on the Daily Mail story (9/9/2009) is on http://amberhawk.typepad.com/amberhawk/2009/09/is-one-adult-in-78-subject-to-surveillance-of-their-communications.html.
Population statistics from Office of National Statistics and Wiki.
Annual Report of the Interception of Communications Commissioner for 2009. http://www.official-documents.gov.uk/document/hc1011/hc03/0341/0341.pdf
Note: Please see the comments associated with the Blog. Evidently the number of warrants signed is not all the warrants associated with national security. This is explained at section 5.1 of the Annual Report in a single sentence. This does not change the analysis substantially.